Preliminary results of survey and limited excavations at Euhesperides (Benghazi) and Tocra are discussed. At Euhesperides part of one residential insula was excavated and a mosaic discovered. Evidence is presented that the insula and the city defences are contemporary, dating from the early fourth to the third centuries BC. The earliest stratified levels beneath the floors of the insula date to the early sixth century BC. The work at Tocra comprised a full survey of all visible archaeological remains. In particular, close study and limited excavation of the Byzantine Fortress suggest that the structure survived long beyond the Arab invasions of AD 642.

In the context of the North African littoral Cyrenaica and Tripolitania appear almost as fertile islands, surrounded by desert on three sides and the Mediterranean to the north (Fig. 1). Between Cyrenaica and Egypt the desert runs to the sea, while between Cyrenaica and Tripolitania lies desert which stretches up to the shores of the Gulf of Sirte. Only to the west of Tripolitania is there a thin coastal strip of cultivable land which runs past the island of Djerba, turning north past Gabes to the productive lands of central Tunisia. As the crow flies only some 350 miles (450 km) separate Berenice (Benghazi), the most westerly of the cities of Cyrenaica from Lepcis Magna, her nearest neighbour among the Tripolitanian cities. While a land-route existed along the north African coast, the destinations it offered were clearly limited. Transport by sea not only offered the opportunity for the most economical long distance movement of bulk commodities such as grain, olive-oil and wine — the staples of the ancient world — but it also presented a greater range of possible destinations.

This summary presents the main results of three seasons of survey and limited excavation work carried out by the Society at Ptolemais in Eastern Libya. The survey concentrated on two major town houses which had been partially excavated by Richard Goodchild (House G and House T), whilst excavation was carried out on the adjacent site known as the North-east Quadrant. All of the sites revealed complex structural histories and it is evident that in this quarter of Ptolemais, at least, life within the town continued well into the Islamic period.

The tribal grouping known as the Laguatan, Leuathae and Lawata in various late Roman and Arabic sources are identified as a powerful confederation of Libyan tribes. The confederation comprised two main types of tribes. On the one hand there were neo-berbers who migrated from the east to the west through the Libyan desert in late antiquity; on the other, there were the original inhabitants of the desert oases, of Cyrenaica and of Tripolitania who formed alliances with the newcomers. The growth of the confederation had a destabilising effect on the Roman frontiers and severe raids were made against the more Romanised areas, notably the territoria of the coastal cities.

Through the sedentary agriculture of the allied Libyans, based on settlements such as Ghirza, and new conquest and exploitation, the Laguatan established an economic and agricultural regime largely independent of Rome. It is inappropriate, therefore, to view the Laguatan simply as camel-riding nomads as has been done in the past, nor was the diffusion of the camel a decisive factor in the timing of the onset of their raids. It is argued, on the contrary, that the camel was present at a much earlier date, that it was mainly used as a pack- and farm animal in pre-Islamic times and that the horse was the main instrument of the Laguatan in warfare and raiding. The Laguatan were the instigators of a Libyan cultural, religious and political revival and their history is of great importance to an understanding of the late Roman and Islamic eras.

Active karst landforms (poljes, dolines, sinkholes, caves and rillenkarren) and hydration and weathering phenomena on anhydrite and limestone are described from the Tripolitanian pre-desert. Rainfall in the study area is considerably less than the 250–300 mm a−1 previously suggested as the minimum precipitation threshold for the development of karst landforms. These findings suggest that the presence of karst land-forms in an arid region cannot necessarily be taken as evidence for higher precipitation levels in the past.

This article describes the interdisciplinary methods developed by the UNESCO Libyan Valleys Survey in connection with the study of ancient agriculture. These methods combine the techniques of settlement archaeology with those of the earth sciences and palaeoeconomics. The interactive nature of the enquiry does not resolve all the questions, but it can help to re-shape those questions and suggest new lines of enquiry. The example of the Wadi Mansur is used here to demonstrate the application of both conventional and new techniques and the contribution this can make to our appreciation of ancient land use in an arid zone. This provisional statement of results describes the geomorphology of the wadi, the settlement archaeology, ceramic dating evidence, palaeoeconomic data (including palynological evidence), wadi wall technology and typology.

Uranium/Thorium (U/Th) and 226Ra dating has been applied to spring deposits in order to determine the timing of the drying up of springs in the Wādī al-Ajāl, Fazzān, Libya. These dates are compared with results of archaeological surveys of irrigation structures and archaeobotanical studies of plant remains to further our understanding of the factors that controlled the introduction of irrigated agriculture, its nature and timing.

Though there are still important gaps in our understanding, the results suggest that by 3,100 ±125 BP surface water was either very scarce or nonexistent at the base of the escarpment, although the water table was probably very near the surface at the base of the Wādī in the vicinity of Jarma. Archaeobotanical evidence indicates that irrigated agriculture was introduced in the early part of the first millennium BC, soon after the desiccation of the springs and presumably in response to the fall in the water table and the increasingly scarce availability of surface water. No irrigation structures have been found that relate to this period, but the most likely scenario involves wells tapping the near surface water in the vicinity of Jarma. The earliest irrigation systems are foggaras which archaeological evidence suggests were probably introduced by the final few centuries BC, and definitely before the fourth century AD. The extensification of agriculture associated with extensive foggara development broadly corresponds with archaeobotanical evidence for an intensification and diversification of agriculture involving the introduction of a farming system utilising both winter and summer crops. There are likely to be strong connections between these two developments and that of the rise of the Garamantes.

Recharge of aquifers in the SW Fezzan dates back to the last millennia of the Late Pleistocene and to the very beginning of the Holocene, with a consistent delay between the mountain area (at the beginning) and the ergs. Since that period and up to the end of the fifth millennium BP, the whole area of the Wadi Tanezzuft and of its basin, including mountain ranges and low lands (ergs and pediments), consisted of a wet savannah, and in the low lands displayed a continuous belt of ponds and lakes. During this period, all the physiographic units of the landscape were exploited by hunters and collectors and later, much more intensively, by pastoral communities. Apart from two minor dry episodes, the whole area dried out around 5000 years BP. The wadi Tanezzuft was substantially reduced in discharge and size, but its main course was still fed with water for about three millennia, and during the third millennium BP it was a green oasis about one hundred kilometres long; and for this reason it was intensively settled by the Late Pastoral communities and later by the Garamantes. The different behaviour of the Tanezzuft valley in comparison with the surrounding areas is to be attributed to the fact that it is the main outlet of the Eastern Tassili hydrographic basin, whose water reserve was recharged during the wet Holocene and reduced, but not exhausted by the onset of aridity at 5000 years BP.

During the first centuries AD the size of the oasis was strongly reduced as a consequence of the final depletion of water reservoirs. However strong wind erosion, which is the present dominant geomorphic process, began in the late medieval period.

The archaeological heritage of Libya is at serious risk from uncontrolled development. The reasons for the situation are discussed, and the attendant difficulties in preventing encroachment on the archaeological zone. The case of Cyrene is not alone, but while the focus here is its particular problems, guidelines for safeguarding Libya's heritage are proposed, while acknowledging current efforts to counteract the state of play.

Excavations by the Department of Antiquities, Benghazi, Garyunis University, Benghazi and the Society for Libyan Studies took place at the site of Euesperides for a total of seven weeks in 1995 and 1996. Work was concentrated on the northern margins of the city, where much new evidence for its topography and development came to light. The discoveries include successive fortification walls (the earlier of which may belong to c. 600 BC), part of a necropolis, and a large extramural kiln complex of the fourth/third centuries. Evidence bearing on the infrastructure, economy, diet and cultural life of this early Cyrenaican city, which was abandoned by the mid third century BC, was also recovered.

This article concerns a series of 20 samples submitted for AMS radiocarbon dating from historic sites in the Libyan Sahara. The sites had been identified initially from remote sensing analysis, then visited on the ground in 2011 and organic samples suitable for a dating programme obtained. With the help of an award from the NERC-AHRC National Radiocarbon Facility, this initial suite of dates has been provided. The results are important in several ways. They demonstrate very clearly that settlement in this hyper-arid desert landscape reached its densest pattern in the late Garamantian era, broadly the fourth-fifth centuries AD. In the Islamic era that followed, though the overall population appears less dense, our dates throw light on several possible phases of settlement renewal.

The paper reports on the sixth season of fieldwork of the Cyrenaican Prehistory Project (CPP) undertaken in September 2012. As in the spring 2012 season, work focussed on the Haua Fteah cave and on studies of materials excavated in previous seasons, with no fieldwork undertaken elsewhere in the Gebel Akhdar. An important discovery, in a sounding excavated below the base of McBurney's 1955 Deep Sounding (Trench S), is of a rockfall or roof collapse conceivably dating to the cold climatic regime of Marine Isotope Stage (MIS) 6 (globally dated to c. 190–130 ka) but more likely the result of a seismic event within MIS 5 (globally dated to c. 130–80 ka). The sediments and associated molluscan fauna in Trench S and in Trench D, a trench being cut down the side of the Deep Sounding, indicate that this part of the cave was at least seasonally waterlogged during the accumulation, probably during MIS 5, of the ~6.5 m of sediment cut through by the Deep Sounding. Evidence for human frequentation of the cave in this period is more or less visible depending on how close the trench area was to standing water as it fluctuated through time. Trench M, the trench being cut down the side of McBurney's Middle Trench, has now reached the depth of the latest Middle Stone Age or Middle Palaeolithic (Levalloiso-Mousterian) industries. The preliminary indications from its excavation are that the transition from the Levalloiso-Mousterian to the blade-based Upper Palaeolithic or Late Stone Age Dabban industry was complex and perhaps protracted, at a time when the climate was oscillating between warm-stage stable environmental conditions and colder and more arid environments. The estimated age of the sediments, c. 50–40 ka, places these oscillations within the earlier part of MIS 3 (globally dated to 60–24 ka), when global climates experienced rapid fluctuations as part of an overall trend to increasing aridity and cold.

Digital image data from the Landsat satellite Multispectral scanner have been analysed using a Dipex Image Processor. The image data were classified by a clustering algorithm to produce a thematic map for trial use in areas of archaeological interest. The archaeological, pedological, geomorphic and geological importance of these new data are discussed in the context of ground data obtained during the UNESCO April 1984 field season.

This report discusses the first results of the botanical analysis of 182 samples from seven Romano-Libyan and two Islamic sites situated in the pre-desert of Tripolitania. Nine different crop plants have been discovered: barley, wheat, peas, lentils, olives, grapes, figs, dates and almonds. The likelihood of local production versus import, and of subsistence farming versus production for a market is discussed. No significant difference was detected between the range of crops found on the earlier open farms and those found on the later gsur.

Libya's archaeological heritage is under serious threat, not only because of recent conflict, but also due to other factors such as urban expansion, agricultural development, natural resource prospection, vandalism, looting and natural deterioration. The Endangered Archaeology in the Middle East and North Africa Project (EAMENA) has developed a database and methodology using remote sensing and other techniques to rapidly document archaeological sites and any disturbances and threats to them in Libya and across the MENA region. This paper will demonstrate this methodology and highlight the various types of disturbances and threats affecting the archaeology of Libya, concentrating on four case studies in different areas of the country, including the coastal plain around Zliten, a section of the Wadi Sofeggin in the pre-desert, and the desert oases of Jufra and Murzuq.

The palaeoanthropological and geomorphological sub-projects of the Desert Migrations Project (DMP) focus on the Pleistocene and early Holocene environment and prehistory of Fazzan so as to assess the timing and extent of hominin and human movement across the Sahara through time. This paper reports on the findings of the 2008 field season, with a focus on the prehistoric evidence along the northern margin of the Ubari sand sea.

The geomorphological record of the area preserves evidence of at least five past episodes of lake formation. The exact chronology of these, as well as the spatial extent of these lakes, remains the focus of further study.

The archaeological record of hominin and human occupation of Fazzan prior to the establishment of the Garamantian civilisation is extraordinarily rich. Between 2007 and 2008, the DMP palaeoanthropological project surveyed thirty-five localities along the northern margin of the Ubari sand sea, recording a range of assemblages spanning all Palaeolithic industries. Most of the archaeological remains found consisted of stone-tools, while grinding stones were comparatively restricted geographically. Mode 1/Oldowan tools were found at two localities, contrasting with the widespread presence of Mode 2/Acheulean, Mode 3/Middle Stone Age and Mode 5/microlithic artefacts. This indicates that, although hominin presence in the area is probably earlier than previously thought, populations were comparatively sparse until the Middle Pleistocene. Twenty-one localities within the Ubari sand sea, as well as seven south of the Messak Settafet were also surveyed between 2007 and 2008. The detailed study of the lithics from these areas will be carried out next year, but preliminary results stress the different nature of the assemblages found within interdune corridors — very low frequency of cores, no Mode 1 and extremely rare Mode 2 lithics (found at a single locality).

The 2009 field season will focus on obtaining further samples of palaeolake sediments for dating, on the evidence of Mode 1 assemblages south of the Messak, as well as on the refining of the archaeological indicators that may distinguish the different phases of hominin and human occupation of Fazzan during the Later Pleistocene and Holocene.

Survey by the DMP Burials and Identity team around the Tāqallit headland in 2009–2010 has revealed in exceptional detail a well-preserved Garamantian landscape, comprising extensive cemeteries, foggara irrigation systems and numerous oasis settlements. However, this remarkable survival of the Garamantian landscape was found in 2010 to be under direct and imminent threat of destruction. This report describes the landscape features recorded and the steps taken to try to preserve the evidence from obliteration in the face of modern agricultural development. Important new information was recorded about the date and furnishing of some key types of Proto-Urban tombs, linking with a refined view of the relationship of these cemeteries to contemporary foggara construction and the creation of pioneer farming settlement in the Tāqallit region. Significant additional details of the foggara systems were recorded through a combination of satellite image interpretation, surface observation and selective descent into rock-cut shafts. The discovery of an unexpected number of ancient settlements and structures of Garamantian date represents another major achievement of the work. The composite picture of the Garamantian landscape encompassing cemeteries, foggaras and settlements is arguably the most complete yet recorded in the FP/DMP work.

A group of 25 new AMS radiocarbon dates for historic-era sites in Fazzan is presented. These provide further confirmation of the construction of numerous fortified villages and castle-like structures (qsur) in two of the main oases belts of Fazzan during the Garamantian period, primarily in the third – sixth centuries AD. Further precision is also provided on the dating of a Garamantian and early Islamic urban centre called Qasr ash-Sharraba and the early modern capital of Fazzan at Murzuq.

The sedimentological properties of samples collected from exposures in wadi floors are reported. These studies emphasise the natural variability of erosional and depositional processes in this environment. Some of the sedimentological changes observed appear to be of archaeological significance. They indicate that in the recent geological past, either increased flooding occurring as a result of natural environmental fluctuations, or irrigation practice, has caused the elevation of saline groundwaters in the wadi floor sediments to levels which might have had deleterious effects on the contemporary agricultural system. Sites at the margins of the wadis are more sensitive to such changes than those more centrally located on the wadi floor. No evidence for significant climatic change at the times of the extensive Romano-Libyan occupation of the pre-desert region has been detected. There is, however, evidence of much wetter periods in the more distant past, which have left a legacy of mature soils on the basalt plateaus, which may have been extensively used in the Romano-Libyan period.

The final season of the three-year programme to survey the central Tripolitanian pre-desert examined the region to the west of the regions of the Wadis Sofeggin and Zem Zem, casting important light on the route between the interior of Libya and the coast during the prehistoric and Roman periods. In particular, a detailed account is given of the archaeological sites of the Bir Scedua Basin and of the Severan fort at Gheriat el-Garbia.